1. Field of the Invention
The disclosure relates generally to route management, and, more particularly to reroute methods and systems that perform backward routing based on destination tree, and methods for generation of backward routing information.
2. Description of the Related Art
Since ARPANET, IP (Internet Protocol) network has equipped with response network recovery ability. That is, once a node or a connection is broken down, the IP network can automatically trigger to update route. Network nodes, such as routers can exchange new routing information via IGP (Interior Gateway Protocol), such that the routing information of respective nodes can be converged into a new and best condition.
Generally, the convergence process of the routing information includes error detection, information propagation, route calculation, and routing table update. The error detection can be triggered from the physical layer/data link layer or the network layer. Conventionally, the error detection triggered from the physical layer/data link layer needs hundreds of milliseconds. In the error detection triggered from the network layer based on RIP (Routing Information Protocol), routing information is exchanged every 30 seconds, and a route is presumed as invalid when the routing information is not updated in 180 seconds. In the error detection triggered from the network layer based on OSPF (Open Shortest Path First), routing information is exchanged every 10 seconds, and a router is presumed as breakdown when no response is received from the router in 40 seconds. Since the routing information must be propagated through the network, the transmission of routing information needs about 10 ms˜100 ms for passing through each router. The synchronization time for network is expanded with the network size. The time for route calculation and routing table update is based on the processing ability of the router, and the corresponding algorithm. Further, the time for route calculation and routing table update is also influenced by the network size and the amount of routing information. Therefore, for general IP networks, the route recovery time when the connection or node is broken down is time-consuming. The network recovery speed cannot fulfill the requirement of real-time applications. For example, the network breakdown time for a network telephone application must be less than 200 ms˜450 ms, and the network breakdown time for a multimedia streaming application must be less than 1 sec˜30 sec.
Additionally, since the IP network is limited by shortest path route and hop-by-hop, most mechanisms of backup route calculation based on area information cannot achieve satisfactory protection results. Further, the protection mechanisms for connection-oriented networks, such as SONET, ATM, MPLS are not suitable for backup route calculation of IP network, since source routing is adopted in the connection-oriented networks, and the backup path can be therefore defined and built in advance. As described, since the IP network is limited by shortest path route and hop-by-hop, only a backup next node can be calculated. The next node of the calculated backup next node then follows the shortest path route rule to transmit packets to a destination node. Therefore, how to avoid a routing loop during rerouting is an important issue in IP network.